Prehospital Management of Traumatic Brain Injury

Shirley I. Stiver, M.D., Ph.D.; Geoffrey T. Manley, M.D., Ph.D.


Neurosurg Focus. 2008;25(4):E5 

In This Article

Prevention of Secondary Injury in the Prehospital and Emergency Setting

Increased ICP, cerebral edema, cerebral dysautoregulation, and alterations in brain metabolism are inherent sequelae of the primary brain injury. Exogenous or iatrogenic events exacerbate these secondary injury processes. Patients with multiple traumas frequently incur injuries that compromise cardiopulmonary status, and they are therefore particularly vulnerable to secondary injury. Secondary insults are common and are independent predictors of poor outcome in patients with TBI.[28,47,102] Depth and duration of hypotension, but not hypoxia, has been shown to trend in a dose-response manner with the 3-month functional Glasgow Outcome Scale score.[6] In contrast, hypoxia and hypotension do not have nearly as profound an effect on outcome in patients with extracranial trauma. The prehospital guidelines were developed by the Brain Trauma Foundation ( to standardize acute TBI care and prevent secondary injuries and insults.[19,59,146] Implementation of these guidelines with prevention and treatment of secondary injury in the early phases of care significantly improves patient outcomes following severe TBI.[160]

Brain Oxygenation

Airway obstruction and aspiration are major causes of death in patients who die of treatable head injuries.[149] Under normal physiological circumstances, the brain is dependent on aerobic metabolism to maintain the high adenosine triphosphate requirements of neuronal function, and energy failure occurs after a few minutes of anaerobic metabolism.[104,138] Apnea, even for just a brief period, accompanies most head injuries.[47,103] In response to hypoxia, a compensatory increase in CBF occurs, but not until the partial pressure of O2 drops to < 50 mm Hg.[50,86,101] Results from the Trauma Coma Data Bank showed that a low PaO2 (≤ 60 mm Hg) occurred in 46% of 717 admissions to the emergency department.[26] Arterial desaturations to this degree, even for a short time, were associated with a 50% mortality rate and 50% severe disability among survivors.[143] Multiple studies have confirmed that a low PaO2 (< 60 mm Hg) correlates with worsened patient outcomes.[23,28,104,156]

Although airway control is intuitively important, a few studies have unexpectedly reported worse outcomes for patients with TBI who were intubated in the field.[14,33,37,41,53,110,158] In the field, hypoxia is defined as apnea, cyanosis, an O2 saturation < 90%, or a PaO2 < 60 mm Hg. Marked degrees of hypoxia to saturations < 70% are common during intubation with 57% of patients experiencing transient hypoxia lasting a mean of 2.3 minutes.[51] The risk of desaturation associated with intubation is dependent on the starting O2 saturation and occurs 100% of the time if the O2 saturation is ≤ 93%, versus 6% if it is > 93% (p < 0.01).[38] A comparison of 1797 patients with severe TBI who were intubated in the prehospital setting with 2301 patients who were intubated in the emergency department demonstrated a 4-fold increase in death and a significantly higher risk of poor neurological and functional outcome in the group of patients who were intubated in the field.[158]

Endotracheal intubation is associated with risks of increased ICP, aspiration, and hypoxia. The detrimental effects of aspiration or hypoxia before arrival of the emergency response team may not be reversible.[38] Positive pressure ventilation can increase intrathoracic pressure, which may decrease venous return, and in a hypovolemic patient this can impair CPP. Furthermore, sedative agents used during RSI can cause hypotension.[39] Prehospital intubation may also delay transport. Importantly, endotracheal intubation in the field predisposes the patient to an increased incidence of overly aggressive hyperventilation that has been shown to adversely affect outcome[34,36,110] A series of 851 patients with TBI, prehospital intubation, and a range of PCO2 levels on arrival to the emergency department (17%, PCO2 < 30; 47%, PCO2 30-39; and 26%, PCO2 > 40) were divided into 2 groups. Those within the target PCO2 range of 30-39 had a mortality rate of 21%, whereas those whose PCO2 persistently remained outside the target range experienced a mortality rate of 34%.[159] With increasing head severity, there was an increased survival benefit for patients who attained their goal PCO2 level.

Airway management and the prevention of hypoxia is a priority.[20] All patients should receive supplemental O2 to maintain saturations > 90%. In the prehospital setting, intubation has been a mainstay procedure in the treatment of patients with severe TBI and GCS scores ≤ 8, both for maintenance of good oxygenation and prevention of aspiration. The prehospital guidelines of the Brain Trauma Foundation recommend that unconscious or unresponsive patients with GCS scores ≤ 8 or those unable to maintain an adequate airway, and those with hypoxemia (arterial O2 saturation < 90%) despite supplemental O2, should be intubated. In the prehospital setting of a comatose patient with a head injury, orotracheal intubation is preferred over nasotracheal intubation because the status of possible basal skull fractures is unknown and noxious stimulation of the nares can elevate ICP. Intubation not only enables adequate oxygenation, but can also alleviate hypercarbia that itself can worsen ICP. In the absence of signs of raised ICP, patients should not undergo hyperventilation prophylactically.

A consensus panel has addressed the question of poorer outcomes in prehospital patients with TBI who have been intubated. This panel found no prospective controlled trials to adequately address the efficacy of paramedic RSI for severe TBI.[35] Concerns were raised regarding the use of a GCS score alone to identify patients who warrant intubation.[43] The GCS score provides no information as to patient oxygenation status and there is poor interobserver reliability for this score.[8,65] Furthermore, the force of the impact is often sufficient to induce a ventilatory pause, the duration of which is proportional to the force, and early assessments of GCS scores immediately after injury may be inaccurate. The consensus panel suggested that other factors and methods such as pulse oximetry and transport time be incorporated, in addition to the GCS score, to define candidates for intubation.[35] Pulse oximetry and capnometry were recommended for monitoring depth of sedation during conscious sedation. Importantly, success of intubation with RSI does not alone improve outcome.[37,42] The continued treatment of the ventilated patient with careful monitoring of end-tidal CO2 and prevention of hyperventilation is paramount to achieving improved outcomes in the field.[36,122]


Numerous studies have shown a significant association between hypotension and poor outcome in patients with head injuries.[16,26,27,56,92,95,111,132,143,156] In a study of 613 patients from the Traumatic Coma Data Bank, a single episode of hypotension (systolic blood pressure < 90 mm Hg) in the field doubled the mortality rate.[27,120] Estimates suggest that 8-13% of patients with severe head injuries are hypotensive at the injury scene or in the emergency department.[26] Isolated head injury does not cause hypotension. The brain's ability to extract O2 protects it from hypoxia as long as cerebral perfusion is maintained.[26] In accordance with this concept, hypotension has been reported to be a stronger predictor of poor outcome than hypoxia.[6,26] Cerebral ischemia is evident in the vast majority of patients who die of head injury.[66] Secondary ischemia is a manifestation of the loss of autoregulation and is more common and more severe with increasing severity of brain injury.[25,95,99] In the absence of intact autoregulation, brain perfusion becomes passively dependent on the systemic blood pressure and hypotension leads to hypoperfusion and brain ischemia. Furthermore, in experimental animal models, mechanical injury decreases the threshold for ischemic damage and neuronal loss.[77] Low CBFs are frequent in the early hours following head injury, and even in the absence of blood loss a brief hypotensive episode can initiate irreversible cell death mechanisms in injured neurons.[116]

Despite the strong association between poor outcome and hypotension in patients with head injuries, the general trauma literature has questioned the value of time spent securing intravenous access and administering fluids in the field.[87,131] It has been argued that obtaining intravenous access is time-consuming and the small volumes infused during short transports may not significantly affect outcome.[139] Arguments in favor of delayed resuscitation with "permissive hypotension" include reports that administration of intravenous fluids to actively bleeding patients, before definitive surgical control, may increase blood loss because of hemodilution, higher blood pressures, impaired thrombus formation, and clot disruption.[12,117] For patients with severe head injuries, delayed fluid resuscitation risks significant secondary injury that occurs when CPP falls.[17,124,125] Patients with head injuries treated with delayed resuscitation have been shown to experience progressive intracerebral swelling and increased ICP, as well as higher lactate/pyruvate ratios as a result of delayed restoration of CBF.[2]

Management of hypotension in the field improves outcome for patients with severe TBI.[64,150] In children, systolic pressure goals are lowered in an age-dependent manner.[83] Intravenous fluids should be administered to avoid hypotension or to minimize the duration and extent of hypotension. The scalp has a rich blood supply and scalp lacerations should be addressed as a treatable cause of hypotension. Excessive bleeding consumes coagulation factors and platelets. Administration of intravenous fluids in the setting of excessive bleeding may further worsen coagulopathy through mechanisms of dilution of clotting factors and platelets, hyperchloremia leading to acidosis, hypocalcemia, and hypothermia.[58]

The optimal fluid for resuscitation has not been clearly determined. Intravenous fluids should be isotonic to reduce brain swelling and cerebral edema. Normal saline is preferred over lactated Ringer solution, and solutions containing 5% dextrose (D5W) should be rigorously avoided. Hypertonic saline expands intravascular volume by 4- to 10-fold that of the volume infused.[49] Hypertonic saline provides the potential benefit of enabling blood pressure stabilization with smaller volumes. It supports CPP without aggravating leakage and extravascular fluid accumulation that leads to cerebral edema and increased ICP, which occurs following infusion of high volumes of isotonic fluids.[48] Hypertonic saline also acts in a fashion similar to mannitol and induces an osmotic diuresis that assists treatment of raised ICP. Resuscitation with hypertonic saline doubles the survival rate of patients with TBI who present with hemorrhagic shock.[151,155] In subgroup analyses, patients with TBI and GCS scores ≤ 8 are most likely to benefit from hypertonic saline.[150,152] However, studies have yet to demonstrate an improvement in long-term neurological outcomes for patients with head injuries resuscitated using hypertonic saline.[31] A prospective intervention trial is planned to compare hypertonic saline, hypertonic saline with dextran 6%, and normal saline in a subgroup of patients with TBI and GCS scores ≤ 8, both with and without hypovolemic shock (systolic blood pressure ≤ 90 mm Hg).[18]

Management of Suspected Raised ICP

The motor component of the GCS score is the most informative measure of the score for long-term outcome.[29,47,97] Hypoxemia and hypotension can lower a patient's GCS score, and thus assessments made following resuscitation are more reliable.[94] If possible, a GCS score prior to administration of sedative or paralytic agents serves as a valuable baseline. Serial examinations are important. A drop of 2 or more points in the GCS score is considered significant and should raise an index of suspicion for an expanding intracranial mass lesion.[135] In a series of 81 patients with initial field GCS scores of 13 or 14 who subsequently deteriorated and required prehospital intubation, 31% had an abnormal CT scan and 21% had evidence of an intracranial hemorrhage.[54] The development of an oval or irregular pupil is the first sign of uncal herniation.[98] Signs of cerebral herniation also include asymmetric pupils with a difference > 1 mm in size, unilateral or bilateral dilated and fixed pupils, extensor (decerebrate) posturing, or a progressive loss of 2 points in GCS score starting from an initial GCS score ≤ 8.

In the setting of signs of cerebral herniation, urgent measures are needed to lower ICP. These measures should include acute hyperventilation and mannitol administration.[154] In addition, good outcomes for patients with traumatic intracerebral mass lesions correlate strongly with prompt surgical evacuation.[72,121,133,144] In the face of a declining GCS score, the prime goal should be definitive neurosurgical care. In these situations, timing becomes critical as outcome directly correlates with the duration of the mass effect.

Mannitol Administration. The mechanisms underlying the therapeutic benefits of mannitol are not thoroughly understood. By increasing the osmotic gradient between blood and the brain, water is drawn from normal and edematous brain into the vascular compartment, leading to prompt osmotic diuresis and a reduction in ICP.[76,161] The onset of action is within 15-30 minutes with a peak response at 1 hour, lasting for 6-8 hours.[5] Mannitol is more effective when given in a high dose (1.4 gm/kg) as a bolus rather than by continuous infusion.[21,32,142] Mannitol also reduces blood viscosity and improves the rheology of blood flow.[24] There is evidence to suggest that mannitol also acts through vasoconstriction in response to these changes in blood viscosity.[106,108]

Hypertonic Saline. Hypertonic saline has been shown to be as effective as mannitol in treating raised ICP in patients with head trauma.[48,57,112] Concentrations ranging from 7.5% (2 mg/Kg) to 23% (1 ampule, 30 ml) are effective.[71,145,154,162] Hypertonic saline acts through osmotic, vasodilatory, hemodynamic, antiinflammatory, and neurochemical mechanisms.[48] Through osmotic effects, hypertonic saline draws fluid into the intravascular compartment, reducing brain water and improving perfusion. In both human and animal studies, hypertonic saline has been shown to increase mean arterial pressure, likely as a result of plasma volume expansion.[48] Problems of volume depletion and hypotension are not as profound with hypertonic saline as they are with mannitol.[71,109,114,153]

Hyperventilation Therapy. Moderate hyperventilation therapy is indicated as a temporizing, life-saving intervention for the comatose patient with impending cerebral herniation. Carbon dioxide is a potent vasodilator of the cerebral microcirculation. For every 1 mm Hg drop in PCO2 there is a concomitant 3% decrease in CBF.[15] Hyperventilation quickly decreases cerebral arteriolar diameter and can dangerously lower CBF.[96,111] Microdialysis and brain tissue O2 measurements have demonstrated that even brief periods of hyperventilation can lead to hypoxia and clinically significant changes in metabolites, together, indicative of cerebral ischemia.[93,96] The potential for hyperventilation to compromise CBF is especially true in the early phases of severe brain injury when autoregulation is impaired and CBF is reduced from the injury itself.[95,107,137] Ventilatory interventions to hyperventilate the patient frequently also alter thoracic pressures and may have deleterious effects by decreasing venous return and raising ICP. For these reasons, hyperventilation should be reserved for situations of acutely increased ICP with impending cerebral herniation.[22]

Without clear evidence of cerebral herniation, pCO2 goal values in the prehospital setting should be in the range of 35 and 40 mm Hg, typically achieved with tidal volumes of 10 ml/kg and ventilatory rates of 10 breaths/minute.[74] Herniating patients should be hyperventilated to PCO2 levels not < 30 mm Hg, with a goal of 30-35 mm Hg. End-tidal PCO2 may not be a reliable measure of PaCO2, particularly in patients with multiple traumas, chest injuries, hypovolemia, and hypotension secondary to massive blood loss.[11,74,134] In practice, studies continue to show a relative disparity between the guidelines and clinical practice.[147,159] Manual ventilation, with high ventilatory assist rates, appears to predispose the patient to low end-tidal CO2 levels.[147]


Hypothermia is a predictor of death in trauma patients.[80,89] A retrospective analysis of 38,520 trauma patients, aged ≥ 16 years with an admission body temperature ≤ 35°C, demonstrated an increased risk of death in all patients (OR 3.03, 95% confidence interval 2.6-3.5) as well as in a subgroup of patients with isolated, severe TBI (OR 2.21, 95% confidence interval 1.6-3.0).[157] In experimental models, induced hypothermia has several physiological actions beneficial to TBI. Hypothermia acts on the central nervous system to decrease metabolism, and for every 1°C drop in temperature there is a 6-7% decrease in the cerebral metabolic rate for O2.[55,126] The results of several good quality meta-analyses showed a benefit to induced therapeutic hypothermia for severe TBI, but failed to demonstrate statistical improvements in long-term clinical outcome.[119] During induced therapeutic hypothermia, shivering and catecholamine responses are controlled and the pathophysiological mechanisms may differ from those occurring in spontaneous exposure hypothermia, wherein stress responses proceed unabated. Further, the duration of induced hypothermia is usually 24-48 hours or longer as compared with the relatively short periods for field hypothermia. Rebound increases in ICP may also occur during rewarming.[13] However, early spontaneous hypothermia, such as that which occurs with field hypothermia, may be effective in minimizing brain injury if patients are not rewarmed. Young patients < 45 years of age who present with hypothermia (< 35°C) experienced a statistically higher incidence of better ICP control and a lower incidence of poor outcome (76%) if they were kept hypothermic (33°C) for 48 hours, as compared with those who were warmed on admission.[30,90]